Interesting Neurals?

[michaelg]

If one had over the past two days bet this system:

1) Minimum of 8 starters,
2) Neural default settings,
3) Minimum of 3 horses under double figures in the resulting neural market,

then the results using Unitab divvies are:

Friday
Outlay of $21
Return of $45.10

Saturday
Outlay of $91
Return of $128.20


There were 36 races and the winner was struck in 24 of them for a strike rate of 66%.

Overall.
Total outlay of $112.00
Total return of $173.30

[jose]

"3) Minimum of 3 horses under double figures in the resulting neural market,"


Maximum???? perhaps MG.
But it looks like something I work on in a slightly different way.

[michaelg]

Thanks, Jose - it should have read "maximum of 3 horses".

I've been doing something similar with a minimum of 4 selections also with the neural default market, but laying the field.

[jose]

Same here basically, but I am using them for GB races oddly enough.

[michaelg]

With three races to go, the Lay the Field method cannot lose.

Jose, you've piqued my interest about applying the method to the U.K. - do the neurals load their scratchings, and if so, when? And I presume you're successful?

The Win method described in the initial listing of this thread had a losing day. The outlay was $44, and the return was $40.20. However the Betfair S.P (return was $51.58. In reality the return would have been even more because I deducted 5% of the winning dividend instead of the gross profit.

From the 15 qualifying races the winner was struck in 10 of them for a 66% strike rate.

[jose]

I actually use the DS market, (don't trust those neurals enough for real $$$) but it is similar.
If there is 5 horses <=$12 in the DS market, I lay the field, but it has become a bit erratic of late.
Today:
9 races (1 to run) for 1 favourite winning. That race was Devonport R1, where incidentally there were 6 favs winning from the 8 races.
Seems to work for the UK races so far. Touch wood.
Cheers.

[michaelg]

Jose, I used to use D.S's market but I too found it somewhat unsuitable for laying the field, mainly because I didn't want any races where the fave was too short - I don't think D.S's market is ideal for this.

It seems the neurals, especially the combination of the CP and CF algorithyms, except horses that are resuming from a spell, are more suitable.

I also see that today's Devenport R1 had only seven starters - I am quite hesitant to lay these smaller races without knowing the prices just prior to the start of the race.

I had an excellent day with Laying The Field based on the neurals, over $300 profit, and this is putting the bets on as soon as the scratchings are loaded, and not having to be at the computer just prior to the race to be aware of the prices.

[The Ocho]

If one had over the past two days bet this system:

1) Minimum of 8 starters,
2) Neural default settings,
3) Minimum of 3 horses under double figures in the resulting neural market,

then the results using Unitab divvies are:

Friday
Outlay of $21
Return of $45.10

Saturday
Outlay of $91
Return of $128.20


There were 36 races and the winner was struck in 24 of them for a strike rate of 66%.

Overall.
Total outlay of $112.00
Total return of $173.30
Hi michaelg. You've piqued my interest in that this could be something I could load onto a bot before work each morning.

Just a few of questions.
1. What are the default settings literally? I've mucked around with them and can't figure out how to get it back. I press default and nothing happens (maybe they are the default settings :oops: )
2. What are your figures based on - $1 outlay per selection?
3. Could it/would it work better if the outlay was placed on winning the same amount for all runners? Say betting to win $5 whatever price the runner is.
4. The 66% strike rate that has come up the last few days is the race strike rate I take it. Obviously betting on 44 horses with 10 winning is not 66%.
5. Have you checked this out with more data than just the last 3 days?

Sorry for all the questions but I might as well get it right if I'm going to give it a go.

[thorns]

Most likely a blonde question from me, but which horses are you backing? The ones in single figures?

[Bhagwan]

Minimum of 3 horses under double figures in the resulting neural market,

it should have read "maximum of 3 horses".

Hi Michael
Is that the Neural Accessed Price market or live BF Market.

Cheers.

[michaelg]

Hi.

The Default settings are all neural categories put on a setting of 3 except WT which is on zero.

The results are based on a level $1 bet on every selection to Win regardless of price, so that there is a maximum of $3 bet on every race.

The single-figured selections are those shown under $10 in the Neural Assessed Price. If there are more than three selections in a race then it does not qualify for the method. Hopefully the neurals are saying that its not an "open" race but with one of few chances.

I think the strike rate isn't too bad because it probably is superior to betting the top three faves per race.

Unfortunately this is from data over the past three days (41 races), and of course upcoming results could drastically change for the worse.

The Ocho, yesterday morning the scratchings at a particular time had not been loaded on to the neurals, so I manually did so after identifying them from the TAB. I then recorded the selections. However when R&S themselves loaded the scratchings to the neurals I noticed there was a discrepancy with the selections and qualifying races - so a word of caution if you manually load the scratchings.

Hopefully I can list here today's selections but I'll have to wait for the scratchings to be recorded by R&S.

[jose]

After last night's dismal effort I think you may be right MG.
Of the 4 races for that method there were 2 wipeouts.
Pass me that drawing board please.

[michaelg]

Jose, maybe if you looked at the neurals as I do instead of D.Scott's market? I checked his market today, the scratchings for last night's U.K. racing were not shown in his market or the neurals - unless there were no scratchings.

Only one race today:

W'bool
2/ 7, 9, 17.

[The Ocho]

Thanks michaelg. I checked last night and there were 4 eligible races but today, following the scratchings not one of those races qualify and, instead, a different race as been selected - which, by the way, lost today.

[michaelg]

Yes, today's only race lost.

Out of interest I've checked the Place results but only for today and yesterday. If I get time I'll check Friday and Saturday's Place results.

On Sunday, Win betting on the QLD TAB lost $3.80 but Place betting won $0.20.

And today the TAB loss for Win betting was $3.00 but for the Place there is a profit of $1.50.

[michaelg]

Unfortunately I did not record the selections for Saturday only the Win results, and as I now go back into Sat's neurals they've all changed.

However, from now I'll record both Win and Place results just in case they both prove worthwhile.

[The Ocho]

No wocking furries michaelg.

It looks like another method I can't use pre-work however good luck with it.

[michaelg]

Two races today:

Pakenham
5/ 4, 5, 7

Scone
1/ 4, 5, 6

[michaelg]

The winner was struck in both races. For an outlay of $6 the return on the NSW TAB was $5.90 (Betfair S. P it was $6.46) and for the Place it was $10.60.

[enjay]

Two races today:

Pakenham
5/ 4, 5, 7

Scone
1/ 4, 5, 6

Hi.
Please tell me where I am going wrong. I have looked at various times during the day and even now, Scone race 1 shows 4 horses under $10.
4 @ $3.10 5 @ $3.90 6 @ $5 and 9 @ $9.00

Enjay

[michaelg]

Enjay, it's possible I made a mistake with the Scone race. However the neurals often change, even after the running of the first race, and I think that's what happened.

[enjay]

Enjay, it's possible I made a mistake with the Scone race. However the neurals often change, even after the running of the first race, and I think that's what happened.

Thanks Michael.

I first looked at about 10am this morning and it was showing 4 horses.

However thanks for the idea, will keep an eye on it.

Regards
Enjay

[michaelg]

Enjay, I think I must have made a mistake. However, the Pakenham race was a good each-way result.

I remember some time ago that someone said a good system was to bet for a Place the shortest-priced horse in the pre-post market at every venue. But it died a slow death.

Out of interest, over the past two days I looked at the highest points for every venue with the neural default settings. There were 5 venues and all selections placed.

On Monday the selections were - the TAB divvies are shown next to them,

Geraldton 6/1 - $1.40
Warnambool 2/17 - $3.20

Tuesday
Scone 8/7 $1.70
Pakenham 5/4 - $1.60
Townsville 2/5 - $2.60

[michaelg]

Three races today. I've omitted Ascot because R&S haven't yet loaded the scratchings.

Wyong
1/ 3, 6, 12

Launceston
2/ 3, 7, 10

Eagle Farm
1/ 1, 9, 11


The top neural points for each venue are:

Ascot 5/2
Eagle Farm 1/1
Gawler 3/5
Launceston 2/3
Sandown 5/3
Wyong 7/7.

Because I'll bet them to Place they'll probably all lose.

[michaelg]

Correction - Sandown 5/3 is scratched. The selection is Sandown 5/4.

[michaelg]

Not a good day. The outlay was $9, the Tab return was $7.90, Betfair S.P was $8.92 and the Place was $6.90.

The losing race of Wyong R1 had half the field having their first career start. If I had omitted the race it would then have been a good day. Aah...hindsight!!!

[michaelg]

Today's races and selections are:

Ballarat
4/ 6, 11, 12

Rocky
4/ 1, 7, 8
5/ 3, 4, 5


I've also been testing a successful similar each-way system with Don Scott's market combined with Unitab's ratings. I've been betting them for the Win with Betfair S.P and for the Place with the NSW TAB. Even quinellas are showing a profit but unfortunately I haven't kept accurate records as the strike rate is low which does not (at the moment) really interest me.

I'll list them and hope the Curse does not get them.

Ballarat
3/ 1, 3, 8
5/ 1, 5, 10
6/ 1, 2, 3, 7
7/ 1, 2, 7
8/ 4, 8, 10

Bordertown
3/ 2, 3, 5

[michaelg]

A good day for the neural system.

From the three races the method struck two winners for a profit of $14.36 with betting the minimum of $5 on Betfair S.P.

Out of interest, if the profit was calculated by deducting 5% from the winner and not the gross profit then the day's profit would have been $12.85 and not $14.36 - this is by my calcs a 11% profit advantage.

Place betting also had a good day. The outlay was $9 and the NSW TAB return was $11.10 for a POT of 23%.


Hopefully the other system (Don Scott and Unitab) will also have a good day. I'll post the results when available.

[michaelg]

A bad day for the Don Scott/Unitab method.

I made a mistake with Ballarat R6. There were four selections which I mistakenly listed, it should have been a no-race. Just like the neural system it becomes a qualifying race only if there are three selections. This seems to the ideal amount for the method. And the Place bet seems to be superior to the Win. If I had omitted Ballarat R7 the outlay would have been $16 for a return of $20.70

[michaelg]

Today's selections for the neural method are (each-way bets):

Ararat
3/ 4, 5, 12
7/ 3, 4, 5

Cranbourne
2/ 6, 7, 8
8/ 2, 4, 8


D.Scott/Unitab system (Place only)

Ararat
2/ 1, 2, 3

Gosford
4/ 8, 12, 13

Ipswich
4/ 2, 5, 8

Orange
1/ 1, 3, 4
7/ 1, 5, 7

[michaelg]

I didn't include Albany in the neural system because the scratchings were taking too long for R&S to record them.

I won't include them in the records, but as a matter of interest these are them:

2/ 1, 2, 4 (the selections ran first, second and third. The winner paid $7.70 ($11.44 with betfair S.P.) and $4.60 for the Place.
5/ 1, 3, 5
7/ 1, 6, 9

[michaelg]

Not a bad day. The Win profit on Betfair S.P. was $22.73, but the TAB Place loss was $0.90.

However if Albany had been included a further $63.74 Betfair profit on Win betting would have been made and also $1.10 on Place betting, which would have given it an overall profit of $0.20.

Out of interest, box quinellas would have cost $12, the Unitab return was $23.30, and with Albany the outlay would have been $9 for a return of $19.50

[michaelg]

Here are today's races and selections. There are a couple of other qualifying races but I've omitted them because of several runners having their first race start.

Warwick Farm
3/ 4, 8, 9
5/ 4, 7, 11
8/ 1, 4, 9

Flemington
1/ 1, 2, 3
5/ 2, 4, 11
6/ 4, 6, 22 (No Hay List ???)
9/ 3, 8, 11

Morphetville
2/ 1, 3, 4

Newcastle
3/ 2, 3, 8
5/ 1, 6, 9

Gold Coast
1/ 1, 2, 6
4/ 1, 4, 6
5/ 1, 2, 13

Queanbeyan
2/ 6, 7, 10
3/ 5, 9, 10
5/ 1, 7, 11
6/ 1, 4, 10
9/ 2, 8, 10

Toowoomba
3/ 2, 3, 5
4/ 1, 2, 7
7/ 1, 2, 4

[michaelg]

Correction. The qualifying race is Newcastle R2, not R3.

[Chrome Prince]

My worry about Hay List is that first up gut busting effort which gave Black Caviar a scare at the 400m mark.
Was probably Hay List's biggest run to date, might have trained on, but I'd be wary.

[michaelg]

Chrome, I hope you're right, but he's so consistent finishing in front of just about the best.

Ascot's selections:

3/ 1, 2, 4
4/ 3, 6, 8
8/ 1, 3, 4

[Chrome Prince]

Well, nearly got nutted, but must have trained on well.
Sorry about that Chief :(